Prevention of Cardiovascular Events with an Antihypertensive Regimen of Amlodipine Adding Perindopril as Required versus Atenolol Adding Bendroflumethiazide as Required, in the ASCOT-BPLA

• p. 6 Results, first paragraph

  18 965 people (99%; figure 1)

• p. 2 Figure 1

  9518 complete information

Judge's note: Figure 1 gives 9518 + 9447 = 18 965 patients with complete information. Out of 19 257 evaluable patients that is 98·48%. Out of 19 342 randomised it would be 98·05%. Neither denominator rounds to 99%. The discrepancy is small.

Second check: another model agreed

• p. 8 Figure 7, 'Age older than 60 years' row

  1001 (17%)

• p. 8 Figure 7, 'Age older than 60 years' row, atenolol

  1186 (20%)

• p. 3 Table 2, age >60

  6081 (63%)

Only the authors can say what is right here.

Judge's note: Every other subgroup row checks against the Table 2 denominators: diabetes 430/2567=17%, 493/2578=19%; smokers 495/3168=16%, 612/3109=20%; female, male and LVH also match; age ≤60 gives 361/3558=10% and 416/3534=12%. Only the >60 row fails, and both values sit just below .5. Only the authors can say whether the subgroup denominators differ or the percentages are mis-rounded.

Second check: another model agreed

• p. 1 Summary, Findings

  Though not significant, compared with the atenolol-based regimen, fewer individuals on the

Judge's note: I checked the Summary on page 1 in the crop. Only the primary endpoint is non-significant (p=0·1052). The other listed outcomes have p≤0·025. Read literally, the sentence puts every outcome under 'not significant'. The printed p values make the intended meaning clear, so this is a wording ambiguity only.

Second check: another model agreed

• p. 5 Results, first paragraph

  characteristics of the 19 257 patients randomised

• p. 2 Figure 1

  19 342 randomised

• p. 5 Results first paragraph

  baseline characteristics of the 19 257 patients randomised

Judge's note: Figure 1 (page 2) gives 19 342 randomised, 85 excluded and 19 257 evaluable. The Results text on page 5 calls the 19 257 'randomised'. The same paragraph goes on to say that 85 patients were excluded, so readers are told about the gap. The label is loose rather than a wrong count.

Second check: another model agreed

• p. 7 Discussion, first paragraph

  terms of reducing the incidence of all types of

• p. 5 Figure 4

  1·27 (0·80–2·00)

• p. 7 Discussion, first paragraph

  terms of reducing the incidence of all types of cardiovascular events and all-cause mortality

• p. 7 Discussion para 1

  in terms of reducing the incidence of all types of

Only the authors can say what is right here.

Judge's note: In Figure 4, five cardiovascular endpoints are non-significant: the primary endpoint (0·90, p=0·1052), heart failure (0·84, p=0·1257), silent MI (1·27, p=0·3089), chronic stable angina (0·98, p=0·8323) and arrhythmias (1·07, p=0·8009). The next sentence of the Discussion itself says the regimen was not significantly more effective for non-fatal MI or fatal CHD. 'All types' is broader than the tabulated results support. The authors could clarify what the phrase is meant to cover.

Second check: another model agreed

• p. 7 Results, last paragraph

  a difference of at least 1%

• p. 9 Table 4 caption

  a difference between treatment groups of more than 1%

• p. 7 Results, last paragraph

  a difference of at least 1% between groups are given in table 4

Judge's note: I compared the Results text on page 7 with the Table 4 caption on page 9 and confirmed the ≥1% vs >1% wording. Every row that is visible differs by more than 1 point, so the table content does not show which rule was used. This is a wording inconsistency only.

Second check: another model agreed

• p. 1 Summary, Methods

  silent myocardial infaction

Judge's note: The page 1 crop shows 'infaction'. It is a misspelling in the published text, not an extraction artefact.

Second check: none (judge only)

• p. 7 Discussion, first paragraph

  not siginificantly more effective

• p. 7 Discussion, first paragraph

  regimen was not siginificantly more effective at reducing

• p. 7 Discussion, first paragraph

  was not siginificantly more effective

• p. 1 Abstract

  silent myocardial infaction

Judge's note: Both are visible in the crops from pages 7 and 12. The other misspellings bundled into this cluster (infaction, generalisibility, Cardiovasular, Amlopidine, inititated) are judged separately under C002, C014, C007, C004 and C003.

Second check: none (judge only)

• p. 2 Methods, Participants

  inititated

Judge's note: The page 2 crop shows 'inititated'.

Second check: none (judge only)

• p. 3 Figure 2 legend

  Amlopidine-based regimen

• p. 4 Figure 3 number at risk

  Amlopidine-based regimen

• p. 6 Figure 5 legend/number at risk

  Amlopidine-based regimen

• p. 7 Figure 6 legend/number at risk

  Amlopidine-based regimen

Judge's note: The embedded figure text layer reads 'Amlopidine-based regimen' at every figure-legend and number-at-risk occurrence (main.txt lines 472, 623, 626, 885–898, 957, 1174–1175). I looked at the crops for Figures 2, 3, 5 and 6 and the legend glyphs match that spelling. The Figure 4 axis label 'Amlodipine-based regimen better' is spelled correctly. Some reviewers said the number-at-risk labels are spelled correctly, but the text layer shows them misspelled too.

Second check: none (judge only)

• p. 5 Figure 4, post-hoc endpoints

  Cardiovasular death+myocardial

Judge's note: The misspelling is visible in the Figure 4 crop on page 5.

Second check: none (judge only)

• p. 9 Discussion

  generalisibility of the overall findings

Judge's note: The misspelling is visible in the page 9 crop.

Second check: none (judge only)

• p. 11 ASCOT Investigators, Sweden

  G R_ter

Only the authors can say what is right here.

Judge's note: The page 11 crop shows the underscore in the printed name, so it is not an extraction artefact. Only the authors can supply the correct letter.

Second check: none (judge only)

• p. 12 Reference 5

  The JNC7 teport.

Judge's note: The page-12 image clearly shows "The JNC7 teport." in reference 5, so this isn't an extraction artefact. It's a typo for 'report'.

Second check: none (judge only)

• p. 12 Acknowledgments

  Drug were

• p. 12 Acknowledgments

  Drug were supplied by Leo Laboratories

Judge's note: The page-12 Acknowledgments crop shows "Drug were" broken across a line and continuing as "supplied by Leo Laboratories". A singular noun with a plural verb is a grammar slip.

Second check: none (judge only)